ROBOT THERAPY FOR FUNCTIONAL RECOVERY OF THE UPPER LIMBS: A PILOT STUDY ON PATIENTS AFTER STROKE

J Rehabil Med 2009; 41: 971 975 ORIGINAL REPORT ROBOT THERAPY FOR FUNCTIONAL RECOVERY OF THE UPPER LIMBS: A PILOT STUDY ON PATIENTS AFTER STROKE Federica Bovolenta, MD 1, Milena Goldoni, PT 1, Pierina Clerici, PT 1, Maurizio Agosti, PT 2 and Marco Franceschini, MD 3 From the 1 Medicine Rehabilitation NOCSAE Hospital AUSL of Modena, Modena, 2 Geriatrics and Rehabilitation Department, University Hospital of Parma, Parma and 3 IRCCS San Raffaele Pisana, Rome, Italy Objective: To verify the possibility of administering robotaided therapy for the upper limbs in patients after stroke; to evaluate patients degree of acceptance and compliance with the treatment; to establish if the treatment has an effect on motor impairment and functional outcome. Design: Quasi-experimental, uncontrolled study. Subjects: Fourteen patients with chronic hemiparesis after stroke. Methods: Patients were treated with a robotic system for the upper limbs (ReoGo TM ; Motorika Medical Ltd, Israel). Subjects performed the following assessment, at the start (T0), at the end of treatment (T1), and at the follow-up performed one month after the end of treatment (T2): Fugl-Meyer test (FM) for upper limbs; strength evaluation; Ashworth scale; visual analogue scale (VAS) for pain; Frenchay Arm test (FAT); Box and Block test (B&B); Functional Independence Measure (FIM TM ); ABILHAND Questionnaire; Timed Up and Go test (TUG); Euro-Quality of Life questionnaire and; a VAS for treatment satisfaction were administered to the subjects. Results: Total scores of FM, B&B, FAT and FIM TM showed a statistically significant improvement from T0 and T1 (FM p < 0.002, B&B p < 0.012, FAT p < 0.023, FIM TM p < 0.007) and from T0 and T2 (FM p < 0.003, B&B p < 0.011, FAT p < 0.024, FIM p < 0.027). No statistically significant differences were found between evaluations at T1 and T2 (FM p < 0.595, B&B p < 0.491, FAT p < 0.317, FIM p < 0.180). Conclusion: The sample was capable of completing the treatment and demonstrated good participant satisfaction. This pilot study led to the finding of a clinical improvement and excellent patient compliance. It can be hypothesized that the results are robot-dependent and that they were learned and then maintained. However, the study is limited in that a control group was not used. As such, it is desirable to continue this study with a control group, as well as by designing a prospective longitudinal randomized controlled trial study. Key words: stroke, upper limbs, robotics. J Rehabil Med 2009; 41: 971 975 Correspondence address: Federica Bovolenta, Medicine Rehabilitation, NOCSAE Hospital AUSL of Modena, via P.Giardini 1355, IT-41100 Modena, Italy. E-mail: federica. bovolenta@libero.it Submitted March 16, 2009; accepted May 19, 2009 INTRODUCTION Stroke is the main cause of disability in industrialized countries, with a significant impact on individual, family, and societal healthcare. As such, any form of treatment that increases the functional recovery of patients after stroke could significantly reduce the physical, emotional, and financial load that this condition carries for the sufferers, their families and society in general (1 3). Therefore, as far as functional recovery is concerned, various longitudinal studies have shown that the upper limbs are involved at the onset of disease in 85% of cases, and that they remain non-functional 6 months after the acute event in 30 66% of cases, while only 5 20% of cases present complete functional recovery (4). Consequently, over the last few years, rehabilitative medicine has encouraged research in an attempt to identify the modalities, time-frames and proper motivations of the rehabilitative intervention: attempting to identify predictive indicators of outcomes in the acute phases (4, 5); and seeking to understand the neurophysiological mechanisms underlying functional recovery (6, 7), in order to plan the most incisive therapeutic interventions. Within the scope of this research, it has emerged that the proposed exercise must be intensive and specific in order for treatment to be effective (3); in addition, treatment must be repetitive, functional and motivating (8, 9), so as to bring about an increase in performance, as well as learning, acquisition and generalization (10). These requirements seem to be satisfied by robotic devices for rehabilitation (11), both in terms of clinical results as well as in terms of positive effects on healthcare costs and increased efficiency. The use of robotic devices for rehabilitation of the upper limbs shows various advantages, including: a large number of patients that can benefit from robotic therapy, due to the flexibility of robotic systems; excellent acceptance of the therapy among patients; and, finally, that the therapy can be performed by the patient under the supervision of a physiotherapist. Recent technological advances have made it possible to develop robotic instruments capable of performing a safe and intensive rehabilitative intervention. Robotic therapy can be developed in different directions in order to reduce motor impairment and increase functional recovery, even in patients affected by moderate/severe impairments, following injuries 2009 The Authors. doi: 10.2340/16501977-0402 Journal Compilation 2009 Foundation of Rehabilitation Information. ISSN 1650-1977

972 F. Bovolenta et al. to the central nervous system, as in the case of stroke. Numerous robotic systems have been developed for sensory-motor rehabilitation of the upper limbs of hemiplegic patients, such as, for example, MIT-MANUS and its commercial version, the InMotion Shoulder-Elbow Robot (12), ARM-Guide (13), MIME (14), BiManu-Track (15), NeReBot (16), and the REO Therapy System (17). The objectives of this study are: (i) to verify the actual possibility of administering the robotic system for the upper limbs; (ii) to evaluate patients degree of acceptance and compliance with the treatment; (iii) to establish if the treatment has an effect on motor impairment and on functional outcome in these patients. MATERIALS AND METHODS Patients This was a quasi-experimental, uncontrolled study. Patients with chronic hemiparesis were recruited for the study. Persons meeting the following criteria were included: (i) patients with motor impairment and consequent disabilities related to the first acute cerebrovascular event; (ii) outpatients at least one month after suspension of specific treatment for the upper limbs, insofar as they had already reached the objectives of the programme designed by the team, without further signs of changes in the motor picture. Patients were excluded who presented: (i) a lesion located in the posterior circulation; (ii) serious cognitive (Mini-Mental State Examination < 24), linguistic, or perceptual deficits; (iii) an absence of control of the trunk in a seated position; (iv) lack of consent to participate in the study; (v) people who stopped treatment for more than 5 consecutive days were considered drop-outs; under this threshold, any lost sessions were recovered. Treatment Each patient underwent a cycle of treatment with a robotic system for the upper limbs (ReoGo TM ; Motorika Medical Ltd, Israel, Fig. 1) (17). This instrument makes it possible to perform a specific treatment for the upper limbs, in particular through the mobilization of the shoulder and elbow joints. The robot makes it possible to execute movements in 3 dimensions and on all spatial planes. The exercises can be performed in various ways: with forearm support, wrist support only, or through a handgrip. Thus, the system makes it possible to perform numerous kinds of exercises, the purpose of which is to reach the objectives on the computer screen connected to it, with visual and audio feedback. Fig. 1. Robotic system for the upper limbs: ReoGo TM (Motorika Medical Ltd, Israel). The movement mode can vary from completely passive to completely active, through varying degrees of intervention that the patient can exert on the robot s arm. Even the width of the movement itself can be modulated on the basis of each subject s unique characteristics. The treatment consisted of a total of 20 sessions lasting 45 min each, 5 days a week, for a total period of 4 weeks; a protocol we designed was used, with exercises that presented a progression for both movement type (i.e. the joints involved, with a proximal-distal progression) and mode of execution of the movement itself, with a progression from passive movement, to assisted movement, to free movement. Forearm support was used during treatment. Patients did not undergo any kind of specific treatment for the upper limbs during treatment or in the preceding and subsequent month. Evaluations Subjects performed the following assessment: Fugl-Meyer test for upper limbs as modified by Lindmark & Hamrin (18, 19); muscle evaluation of 10 muscles, according to Medical Research Council (MRC) criteria (20); Ashworth scale for spasticity (21); visual analogue scale (VAS) for upper limb pain; Frenchay Arm test (22); Box and Block test (23); FIM motor (24, 25); and the ABILHAND questionnaire (26). In addition, subjects underwent a comprehensive evaluation using the Timed Up and Go test (27). Finally, the Euro-Quality of Life (QoL) questionnaire (28) and a VAS for treatment satisfaction were administered. Evaluations were administered at the start of treatment (T0), at the end of treatment (T1), and at the follow-up performed one month after the end of treatment (T2), during which the patient did not undergo any kind of specific rehabilitation for the upper limb. Statistical analysis A descriptive analysis of the distribution of patients was used to process the data; the following tests were used to verify the existence of a possible relationship between the variables examined: Student s t-test, verified with the Wilcoxon signed-rank test and the exact tests. RESULTS Fourteen patients participated in the study (9 men and 5 women, mean age 60.57 years (standard deviation (SD) 8.18, range 35 71 years)) (Table I). There were 9 cases of ischaemic stroke and 5 of haemorrhagic stroke; 6 of right-side hemiparesis and 8 of left-side hemiparesis; distance from the acute event ranged from 4 months to 25 years; and distance from the last treatment period ranged from a minimum of 30 days to a maximum of 6 months. Only one patient left the study, due to an inability to maintain a seated position for a long time owing to the flare-up of a degenerative disease in a vertebral lumbar disc. One other patient did not attend the follow-up. The Fugl-Meyer test ranged from a total score of 76 (T0) to a score of 85.2 at T1; the improvement was statistically significant (p < 0.002). At T2 the value increased by an additional 2.2 points and was significant (p < 0.003) compared with the value at T0. The Box and Block showed a change from T0, where the mean value was 13.1, to T1, e.g. an increase of 3.9 points, and from T0 to T2, an increase of 6.6 points. The mean value of the evaluations at T1 and T2 was statistically significant (p < 0.012 at T1; p < 0.011 at T2) compared with T0. The Frenchay Arm test, which recorded an average value at T0 of 2.6, obtained an improvement at T1, reaching 3.2 points, and a further increase at T2, reaching a mean

Robot therapy for upper limb recovery after stroke 973 Table I. Description of the sample n (%) Mean (SD) Min max Age, years 14 60.57 (8.18) 45 71 Gender 14 Male 9 (64.3) Female 5 (35.7) Affected side 14 Left 6 (42.9) Right 8 (57.1) Time since stroke (months) 13 49.76 (89.49) 3 291 Disease severity (FMUL) 13 Low 7 (53.8) Moderate 4 (30.7) Severe 2 (15.5) n: sample evaluated; SD: standard deviation; FMUL: disease severity based on the score of the Fugl-Meyer for Upper Limb (modified by Lindmark & Hamrin (19)) at T0: Low FM, 0 35; Moderate FM, 36 75; Severe FM, 76 115. and in the mean value of the evaluations at T2 compared with the initial ones (p < 0.046). The VAS of pain at T0 had a total mean value of 29.8, which decreased notably at T1, to 14.0, and again at T2, where it reached a value of 3.8. The decrease from T0 to T2 was statistically significant (p < 0.010). The Timed Up and Go Test showed a decrease from T0 to T1, with the mean value changing from 18.9 to 18.7. This value decreased further from T1 to T2, reaching 17.3. The decrease from T0 to T2 was statistically significant (p < 0.040). The ABILHAND questionnaire showed an increase from T0 to T1, with the average value changing from 22.4 to 23.6, but there was not a statistically significant difference (p = 0.136). The Euro-QoL had a mean value at T0 of 0.6 and showed an increase at 0.7 at T2, there was not a statistically significant difference (p = 0.229). The VAS for treatment satisfaction had a mean value at T1 of 98.2 (SD 4.01) mm (range 85 100) (Table II). score of 3.6. A comparison of the average values from T0 to T1 was significant (p < 0.023), as was a comparison of mean values between T0 and T2 (p < 0.024). The data recorded by the FIM TM showed a value of 80.1 at T0, which increased at T1 to a statistically significant (p < 0.003) score of 82.6, with an additional increase at T2, where the value corresponded to 84. The same average value at T2 was significant compared with the T0 value (p < 0.027) (Fig. 2). The Ashworth elbow scale had an average value at T0 of 1.7 and showed a decrease of 0.3 at T1 compared with T0, and a decrease of 0.2 at T2. There was a statistically significant difference in the mean value at T1 compared with T0 (p < 0.025), DISCUSSION This study showed that the patient sample was capable of completing the treatment and demonstrated good participant satisfaction. Furthermore, the response of the therapists involved was positive, both from an organizational point of view and with regard to the clinical-rehabilitative responses obtained. Finally, the pilot study showed a clinical improvement for the subjects who took part in it. First of all, an improvement from T0 to T2 was observed on the evaluation scales administered, both in terms of impairment and functionality. At the end of treatment (T1) with robotic therapy, statistically significant changes were observed compared with the initial evaluations Fig. 2. Data at T0, T1 and T2 of total scores of: (a) Fugl-Meyer, (b) Box and Block, (c) Frenchay Arm test, and (d) FIM TM. All showed a statistically significant improvement from T0 and T1 and from T0 and T2. There was no statistically significant improvement between evaluations at T1 and T2. Box-and-whisker Medians (error bars: 95% confidence interval for median).

974 F. Bovolenta et al. Table II. Results for assessment at T0, T1 and T2 T0 T1 T2 n Mean (SD) Min max n Mean (SD) Min max n Mean (SD) Min max VAS for pain 14 29.79 (26.54) 0 80 13 15.77 (23.17) 0 75 12 3.75 (0.62)** 3 5 Timed Up and Go test 14 18.93 (8.53) 10 40 13 18.69 (8.73) 10 38 12 17.33 (7.66)** 9 34 Ashworth shoulder 14 0.36 (0.50) 0 1 13 0.15 (0.38) 0 1 12 0.80 (0.29) 0 1 Ashworth elbow 14 1.71 (1.07) 0 3 13 1.46 (0.97)* 0 3 12 1.50 (1.00)** 0 3 Ashworth wrist 14 0.50 (0.52) 0 1 13 0.54 (0.52) 0 1 12 0.50 (0.52) 0 1 MRC trapezius 14 3.43 (0.51) 3 4 13 3.62 (0.65) 3 5 12 3.75 (0.62)** 3 5 MRC deltoid 14 3.79 (0.58) 2 4 13 4.46 (0.66)* 3 5 12 4.67 (0.65)** 3 5 MRC pectoral 14 3.71 (1.33) 0 5 13 4.54 (0.88)* 2 5 12 4.83 (0.39)** 4 5 MRC internal rotators 14 3.50 (1.16) 0 5 13 3.92 (1.32) 0 5 12 4.08 (1.44) 0 5 MRC external rotators 14 3.29 (1.14) 0 4 13 3.15 (1.63) 0 5 12 3.50 (1.38) 0 5 MRC biceps 14 3.93 (1.07) 2 5 13 4.54 (0.66)* 3 5 12 4.75 (0.45)** 4 5 MRC triceps 14 3.71 (1.38) 0 5 13 4.08 (1.19) 1 5 12 4.25 (1.22)** 1 5 MRC wrist flexors 14 2.93 (1.54) 0 5 13 3.00 (1.87) 0 5 12 3.33 (1.72) 0 5 MRC wrist extensors 14 2.86 (1.46) 0 4 13 3.00 (1.68) 0 5 12 3.33 (1.50)** 0 5 MRC latissimus dorsi 14 2.64 (1.22) 0 4 13 3.15 (1.21) 1 5 12 3.67 (1.07)** 1 5 VAS for satisfaction 13 98.84 (4.00) 85 100 ABILHAND 13 22.38 (9.58) 5 40 13 23.62 (10.15) 7 41 *Statistically significant improvement from T0 and T1; **statistically significant improvement from T0 and T2. VAS: visual analogue scale; MRC: strength evaluation measured with the criteria of Medical Research Council; T0: start of treatment; T1: end of treatment; T2: follow-up (one month after the end of treatment); n: sample evaluated; SD: standard deviation. (T0) on all scales except for the pain evaluation (VAS), the Timed Up and Go Test, the shoulder and wrist Ashworth scale, and the MRC scale of the trapezius muscle, external and internal rotators, triceps, pectoralis major, wrist flexors and wrist extensors. These data indicate an effective improvement in motor performance after administering robotic treatment, even in those patients classified as chronic, i.e. stabilized from a rehabilitation point of view; moreover, these results had been found previously by other studies in the literature (11, 29), thus supporting our own. However, this statistically significant finding was not observed in the analysis of the results between T1 and T2. This fact is interesting, because neither additional spontaneous recoveries nor worsening were found in the time interval during which the patient did not undergo exercise with the robot system; thus, it can be hypothesized that the results are robot-dependent and that they were learned and then maintained. Another very important consideration emerges from the comparison between the evaluations of T0 and T2, where other significances emerge in addition to those found at T1, such as the VAS for pain, certain muscular components (trapezius and wrist extensors) in the MRC and the Timed Up and Go test, which leads one to hypothesize that the maintenance of motor performances of the upper limbs could also improve ambulatory function. This could indicate that statistical significance is also maintained after one month, a period in which the patient does not perform any kind of treatment specific to the upper limbs. The fact that results are maintained is confirmed by other authors at 6 months and at 3 months (12, 30, 31). The fact that the motor performances acquired are maintained leads one to think that the therapeutic training was translated into motor learning. This factor is in line with and Gordon (32); indeed, the robotic instrument makes it possible to administer, in accordance with the theories of Constraint Induced Therapy (33, 34), an intensive, diversified and stimulating exercise that results in changes at the motor and cerebral neuroimaging level (3, 9, 10). This pilot study led to the finding of a clinical increase and excellent patient compliance. However, the study is limited, in that a control group was not used. Despite this, nearly all the patients were known by our centre and had been suspended from treatment, since they did not show changes in the scales that were used in part by our study. In addition, the fact that there were no statistically significant changes between T1 and T2 in either an improving or worsening direction encourages us to undertake further research. Indeed, the former indicates that no spontaneous improvement of the motor and functional picture occurred, and the latter indicates that the improvement recorded was not strictly robotdependent, but rather a sign of motor learning. As such, it is in any case desirable to continue this study with a control group, as well as by designing a prospective longitudinal randomized controlled trial, perhaps focussing on the early stages of inpatient rehabilitation. REFERENCES 1. Pratesi L, Paolucci S, Albuzza M, Franceschini M. Exploring the use of Protocollo di Minima per l Ictus (PMIC Minimal Protocol for Stroke) in two Italian rehabilitation units. Eur J Phys Rehabil Med 2008; 44: 271 275. 2. SPREAD Stroke Prevention and Educational Awareness Diffusion. Ictus cerebrale: linee guida italiane di prevenzione e trattamento [Stroke: Italian guidelines for prevention and tratment]. Milano: Pubblicazioni Catel Hyperphar Group SpA; 2007 (in Italian). 3. Kwakkel G, van Peppen R, Wagenaar RC, Wood Dauphinee S, Richards C, Ashburn A, et al. Effects of augmented exercise therapy time after stroke. A meta-analysis. Stroke 2004; 35: 2529 2536.

Robot therapy for upper limb recovery after stroke 975 4. Kwakkel G, Kollen BJ, van der Grond J, Prevo AJH. Probability of regaining dexterity in the flaccid upper limb. Impact of severity of paresis and time since onset in acute stroke. Stroke 2003; 34: 2181-86. 5. Wagner JM, Lang CE, Sahrmann SA, Edwards DF, Dromerick AW. Sensorimotor impairments and reaching performance in subjects with poststroke hemiparesis during the first few months of recovery. Phys Ther 2007; 87: 751 765. 6. Nudo RJ. Postinfarct cortical plasticity and behavioral recovery. Stroke 2007; 38: 840-845. 7. Prabhakaran S, Zarahn E, Riley C, Speizer A, Chong JY, Lazar RM, et al. Inter-individual variability in the capacity for motor recovery after ischemic stroke. Neurorehabil Neural Repair 2008; 22: 64 71. 8. Yavuzer G, Senel A, Atay MB, Stam HJ. Playstation eyetoy games improve upper extremity-related motor functioning in subacute stroke: a randomized controlled clinical trial. Eur J Phys Rehabil Med 2008; 44: 237 244. 9. van Peppen RP, Kwakkel G, Wood-Dauphinee S, Hendriks HJ, van der Wees PJ, Dekker J. The impact of physical therapy on functional outcomes after stroke: what s the evidence? Clin Rehabil 2004; 18: 833 862. 10. Krakauer J. Motor learning: its relevance to stroke recovery and neurorehabilitation. Curr Opin Neurol 2006; 19: 84 90. 11. Kwakkel G, Kollen BJ, Krebs HI. Effects of robot-assisted therapy on upper limb recovery after stroke: a systematic review. Neurorehabil Neural Repair 2008; 22: 111 121. 12. Volpe BT, Lynch D, Rykman-Berland A, Ferraro M, Galgano M Hogan N, et al. Intensive sensorimotor arm training mediated by therapist or robot improves hemiparesis in patients with chronic stroke. Neurorehabil Neural Repair 2008; 22: 305 310. 13. Kahn LE, Zygman ML, Rymer WZ, Reinkensmeyer DJ. Robotassisted reaching exercise promotes arm movement recovery in chronic hemiparetic stroke: a randomized controlled pilot study. J Neuroeng Rehabil 2006; 3: 12 24. 14. Lum P, Burgar CG, Van der Loos M, Shor PC, Majmundar M, Yap R. MIME robotic device for upper-limb neurorehabilitation in subacute stroke subjects: a follow-up study. J Rehabil Res Dev 2006; 43: 631 642. 15. Hesse S, Schmidt H, Werner C, Rybski C, Puzich C, Bardeleben A. A new mechanical arm trainer to intensify the upper limb rehabilitation after stroke: design, concept and first case series. Eura Medicophys 2007; 43: 463 468. 16. Rosati G, Gallina P, Masiero S. Design, implementation and clinical tests of a wire-based robot for neurorehabilitation. IEEE Trans Neural Syst Rehabil Eng 2007; 15: 560 569. 17. Treger I, Faran S, Ring H. Robot-assisted therapy for neuromuscular training of sub-acute stroke patients. A feasibility study. Eur J Rehabil Med 2008; 44: 431 435. 18. Fugl-Meyer AR, Jääskö L, Leyman I, Olsson S, Steglind S. The post-stroke hemiplegic patient. 1. A method for evaluation of physical performance. Scand J Rehabil Med 1975; 7: 13 31. 19. Lindmark B, Hamrin E. Evaluation of functional capacity after stroke as a basis for active intervention. Validation of a modified chart for motor capacity assessment. Scand J Rehabil Med 1988; 20: 111 115. 20. Medical Research Council. Aids to the examination of the peripheral nervous system. Memorandum 45. London: Her Majesty s Stationery Office; 1976. 21. Ashworth B. preliminary trial of carisprodol in multiple sclerosis. Practitioner 1964; 192: 540 542. 22. Heller A, Wade DT, Wood VA, Sunderland A, Hewer RL, Ward E. Arm function after stroke: measurement and recovery over the first three months. J Neurol Neurosurg Psychiatry 1987; 50: 714 719. 23. Mathhiowetz V, Volland G, Kashman N, Weber K. Adult norms for the box and block test for manual dexterity. Am J Occup Ther 1985; 39: 386 391. 24. Tesio L, Granger CV, Perucca L, Franchignoni FP, Battaglia MA, Russel CF. The FIM instrument in the United States and Italy: a comparative study. Am J Phys Med Rehabil 2002; 81: 168 176. 25. Functional Indipendent Measure: versione italiana. Manuale d uso [Functional Independent Measure: Italian version. Users manual]. Ricerca Riabil 1992; Suppl 2: 1 44 (in Italian). 26. Penta M, Tesio L, Arnould C, Zancan A, Thonnard JL. The ABIL- HAND questionnaire as a measure of manual ability in chronic stroke patients: Rasch-based validation and relationship to upper limb impairment. Stroke 2001; 32: 1627 1634. 27. Podsiadlo D, Richardson S. The timed Up & Go : a test of basic functional mobility for frail elderly persons. J Am Geriatr Soc 1991; 39: 142 148. 28. Dorman PJ, Waddell F, Slattery J, Dennis M, Sandercock P. Is the Euroqol a valid measure of health-related quality of life after stroke? Stroke 1997; 28: 1876 1882. 29. Dipietro L, Krebs HI, Fasoli SE, Volpe BT, Stein J, Becver C, Hogan N. Changing motor synergies in chronic stroke. J Neurophysiol 2007; 98: 757 768. 30. Cordo P, Lutsep H, Cordo L, Wright W, Cacciatore T, Skoss R. Assisted Movement with Enhanced Sensation (AMES): coupling motor and sensory to remediate motor deficits in chronic stroke patients. Neurorehabil Neural Repair 2009; 23: 67 77. 31. Hesse S, Werner MA, Pohl M, Rueckriem PT, Mehrholz PT, Lingnau MA. Computerized arm training improves the motor control of the severely affected arm after stroke. A single-blinded randomized trial in two centers. Stroke 2005; 36: 1960 1966. 32. Gordon J. Assumption underlying physical therapy intervention. Theoretical and historicalperspectives. In: Carr JH, Shepherd PB, editors. Movement science: Foundation for physical therapy in rehabilitation. 2nd ed. Rockville: Aspen Publisher Inc.; 1987, p. 1 31. 33. Dursun N, Dursun E, Sade I, Çekmece Ç. Constraint induced movement therapy: efficacy in a Turkish stroke patient population and evaluation by a new outcome measurement tool. Eur J Rehabil Med 2008 [Epub ahead of print]. 34. Gauthier LV, Taub E, Perkins C, Ortmann M, Mark VW, Uswatte G. Remodelling the brain. Plastic structural brain changes produced by different motor therapies after stroke. Stroke 2008; 39: 1520 1525.